Unreported Alcohol Use Was Common But Did Not Impact Hepatitis C Cure in HIV-Infected Persons Who Use Drugs

Abstract

Background:

We investigated the prevalence and impact of heavy alcohol use on the hepatitis C virus (HCV) care continuum among HIV/HCV co-infected persons who use drugs.

Methods:

In the CHAMPS study, 144 HIV/HCV co-infected persons were randomized to contingent cash incentives, peer-mentors, and usual care to evaluate the impact on HCV care. Alcohol use was ascertained using the 10-item AUDIT (hazardous: male ≥ 8, female ≥ 4) and phosphatidylethanol (PEth) (heavy: ≥ 50 ng/ml), an alcohol biomarker. Log binomial regression was used to evaluate the association between heavy alcohol use and failure to initiate treatment and to achieve sustained virologic response (SVR).

Results:

Of the 135 participants with PEth data, median age was 55 years, 59% were male, 92% were Black, 91% reported a history of drug use, and 97% were on antiretroviral therapy. Hazardous drinking was reported on AUDIT by 28% of participants and 35% had heavy alcohol use by PEth. Of the 47 individuals with a PEth ≥ 50 ng/ml, 23 (49%) reported no or minimal alcohol use by AUDIT. HCV treatment was initiated in 103 of 135 participants and SVR was achieved in 92%. PEth ≥ 50 ng/ml (Relative Risk [RR] 0.72, 95% CI 0.35–1.48) was not significantly associated with failure to initiate HCV treatment or failure to achieve SVR (RR 0.85, 95% CI 0.46–1.57).

Conclusions:

Alcohol use was common and frequently not detected by self-report. However, heavy alcohol use, even when measured objectively, was not associated with failure to initiate HCV treatment or to achieve cure.

Introduction

Among persons with HIV (PWH) in the United States, approximately 25% are co-infected with hepatitis C virus (HCV).¹ Furthermore, about 80% of PWH who inject drugs also have HCV.¹ While antiretroviral therapy (ART) has led to a decrease in opportunistic infections, HCV has emerged as an important contributor to morbidity and mortality amongst PWH. In patients with chronic HCV infection, HIV co-infection significantly increases the rate of liver disease progression and more than triples the risk of liver failure and liver related death from HCV.^1–3 However, the emergence of direct-acting antivirals (DAAs) for the treatment and cure of HCV represents a mechanism to improve the morbidity and mortality of HIV/HCV co-infected patients. In fact, among patients presenting for care in clinical trials and practice for treatment with oral DAAs, HCV cure has been achieved in more than 95% of individuals.^4–6 Despite this success in HCV treatment, significant barriers to cure remain and ongoing alcohol use may be a barrier to progression through the HCV care continuum.

Prior to the DAA era, alcohol cessation was generally required prior to the initiation of HCV treatment as alcohol use could impair immunological responses and reduce the effectiveness of interferon.^7,8 Furthermore, clinicians worried about the potential for ongoing alcohol use to reduce HCV treatment adherence.^7–12 Current HCV guidelines from the American Association for the Study of Liver Diseases (AASLD)/Infectious Disease Society of America (IDSA) prioritize treatment with DAAs for all persons including those with heavy alcohol use.¹³ The paradigm shift around HCV and alcohol use recognizes that DAAs represent short-duration (8 or 12 week courses for the majority of patients), non-toxic (minimal side effects), highly effective, all oral regimens which have the potential through cure to stop progression of liver fibrosis related to HCV. Yet alcohol cessation is often still being mandated by clinicians prior to DAA therapy due to ongoing concerns about treatment adherence and the effectiveness of treatment.^14,15 Furthermore, some state Medicaid plans across the United States restrict access to HCV treatment based on alcohol use.¹⁶ Withholding HCV treatment among HIV/HCV co-infected individuals with ongoing alcohol use is particularly concerning as these individuals are at increased risk of liver fibrosis and liver-related mortality due to the synergistic effects of HIV, HCV, and alcohol use.^2,3,7,17 HCV treatment offers the potential to alleviate an insult to the liver and to also engage patients on topics such as risk reduction and alcohol use. Therefore, we evaluated the prevalence of heavy alcohol use using the Alcohol Use Disorders Identification Test (AUDIT), and the alcohol biomarker, phosphatidylethanol (PEth), and the impact of alcohol use on engagement across the HCV care continuum in a population of HIV/HCV co-infected persons who use drugs who had unfettered access to HCV treatment.

Materials and Methods

The CHAMPS (Chronic HepAtitis C Management to ImProve OutcomeS) study was a randomized-controlled 3-arm trial to evaluate the impact of two strategies, contingent cash incentives and peer mentors, compared to usual care on HCV treatment uptake and sustained virologic response (SVR)/cure as previously described.¹⁸ HCV treatment-naïve persons who use drugs with genotype 1 infection were eligible for the CHAMPS trial if they received HIV care at Johns Hopkins and had not been evaluated for HCV care within 8 months, had a CD4 count >100 mm³, and no evidence of decompensated liver disease. Additional inclusion criteria included: Age ≥18 years, estimated glomerular filtration rate ≥30 mL/min/1.73 m², no evidence of hepatocellular carcinoma, and life expectancy greater than two years. A total of 144 eligible participants were randomized (2:3:3) to 1) Usual Care (UC, treatment with nurse supervision, n=36); 2) UC + Peer Mentors (n=54); 3) UC + Incentives (n=54) from August 2015 to October 2016. All participants were provided 12 weeks of ledipasvir/sofosbuvir (LDV/SOF) at no cost. All participants were linked to an HCV clinician with whom clinical care visits occurred approximately every six weeks along with the clinic-based usual care model of frequent interactions with a nurse-led multidisciplinary team of pharmacists and case managers. Participants also had study-specific visits every six weeks for which they received remuneration ($10-$30 per visit). The primary study endpoint was initiation of LDV/SOF within 8 weeks of enrollment (12 weeks if a change in HIV regimen was required). The study was approved by the Johns Hopkins University School of Medicine Institutional Review Board. All study participants, including peer mentors, provided written informed consent.

Within the CHAMPS trial, we sought to determine the relationship between alcohol use at the baseline visit and two outcomes: failure to initiate treatment and failure to achieve cure. Demographic and behavioral characteristics were collected at the baseline visit through interviewer-administered questionnaires and included the AUDIT to ascertain alcohol use in the preceding 12-month period and the Center for Epidemiologic Studies Depression Scale (CES-D), which produces a score range from 0 to 60, to assess depressive symptoms.^19–21 Clinical measurements completed at baseline included blood phosphatidylethanol (PEth) (USDTL, Des Plaines, Illinois), an alcohol biomarker able to detect alcohol consumption in the past 2–3 weeks, quantitative HCV RNA (COBAS TaqMan HCV Test v2.0; Roche Molecular Systems Inc., Branchburg, New Jersey), CD4 cell count, HIV RNA, liver elastography (FibroScan® 502 Touch, Echosens™ North America, Waltham, Massachusetts), and urine toxicology.^22,23 HCV RNA was also assessed at post-treatment week 12 for participants who initiated treatment to evaluate SVR/cure and 24 weeks after baseline for patients who did not initiate treatment.

Outcomes and definitions

The primary and secondary outcomes of the CHAMPS study, and for this analysis, were failure to initiate LDV/SOF within 8 or 12 weeks (if changes in ART were required) of randomization and failure to achieve SVR/cure, which included both persons who failed to initiate treatment and persons who initiated treatment but did not achieve SVR, respectively. SVR was defined as an undetectable HCV RNA (<15 IU/mL) at 12 or more weeks after stopping LDV/SOF.

Alcohol use was assessed by both AUDIT and PEth. An AUDIT score of ≥ 8 in men or ≥ 4 in women indicated heavy/hazardous alcohol consumption.²¹ PEth ≥ 50 ng/ml was considered positive as this threshold is considered highly sensitive (93%) and reasonably specific (83%) for detecting heavy alcohol use.²⁴ For the liver elastography, participants were classified as either having no/mild stiffness (liver stiffness measurement ≤8 kPa), moderate stiffness (8.1–11.9 kPa) or severe stiffness/cirrhosis (>12.0 kPa) based on standard cutoffs.² A CES-D score of 16 and above was suggestive of depressive symptoms.²⁰ Undetectable HIV viral load was defined as <50 copies/mL.

Statistical Analysis

Baseline demographic, behavioral, and clinical characteristics of the study population were compared by PEth category using chi-square tests for categorical variables and Wilcoxon rank sum tests for continuous variables. Log binomial regression was used to evaluate correlates of heavy alcohol use at baseline. Baseline covariates known or found to be significantly correlated with heavy alcohol use were considered for inclusion in a multivariable model. Prevalence ratios and 95% confidence intervals are reported. Log binomial regression was also used to evaluate the association between heavy alcohol use and failure to initiate HCV treatment and failure to achieve cure. For the outcome of failure to achieve cure, participants failing to initiate treatment were considered to have failed to achieve cure. Baseline covariates known to be or found to be significantly correlated with PEth or significantly associated (p<0.05) with the primary/secondary outcome in the univariable models were considered for inclusion in a multivariable model. Relative risks and 95% confidence intervals are reported. A sensitivity analysis was conducted for the secondary outcome of failure to achieve cure defined only among the subset of participants initiating treatment. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, North Carolina).

Results

A total of 135 of the 144 (94%) HIV/HCV co-infected persons enrolled in the CHAMPS study had PEth data available and were randomized to one of three groups: usual care (n=33), peer mentors (n=52), and cash incentives (n=50) in the CHAMPS study. The median age was 55 years, 59% were male, 92% were Black, 86% were unemployed, 100% were insured, 97% were on antiretroviral therapy, and 12% had cirrhosis (Table 1). Notably, 91% reported a history of drug use while 33% and 26% had a urine toxicology positive for cocaine and heroin at baseline, respectively. By AUDIT, 38 (28%) reported hazardous drinking and 47 (35%) had a PEth ≥ 50 ng/ml which denotes heavy alcohol use. Considering either AUDIT or PEth as an indicator of heavy alcohol use, 61 (45%) individuals met criteria. Of the 47 individuals with a PEth ≥ 50 ng/ml, 24 (51%) had an AUDIT corresponding to heavy/hazardous drinking while 23 (49%) reported no or minimal alcohol use by AUDIT (drinking never, monthly or less, or 2–4 times per month). Of the 88 individuals with a PEth < 50 ng/ml, 14 (16%) reported heavy/hazardous alcohol consumption by AUDIT.

Table 1.

DEMOGRAPHIC, BEHAVIORAL, AND CLINICAL CHARACTERISTICS OF PARTICIPANTS WITH CHRONIC HCV INFECTION BY PETH AT BASELINE VISIT

Characteristic	Total N=135* n (%)	PEth < 50 ng/ml N=88 n (%)	PEth ≥ 50 ng/ml N=47 n (%)	p-value
Intervention Group				0.18
UC	33 (24)	18 (20)	15 (32)
UC+Peer	52 (39)	33 (38)	19 (40)
UC+Incentives	50 (37)	37 (42)	13 (28)
Age, median years (IQR)	55 (50, 59)	55 (50, 60)	54 (51, 57)	0.33
Male	80 (59)	49 (56)	31 (66)	0.25
Black	124 (92)	78 (89)	46 (98)	0.06
Unemployed	116 (86)	71 (81)	45 (96)	0.02
Insured	135 (100)	88 (100)	47 (100)
Urine Toxicology (n=129)
Cocaine	43 (33)	21 (25)	22 (49)	0.006
Heroin	34 (26)	19 (23)	15 (33)	0.19
Recent Cocaine or Heroin Use (self-report) in past 30 days	34 (25)	18 (20)	16 (34)	0.08
AUDIT Score (male n=80, female n=55)
Male <8	62 (78)	43 (88)	19 (61)	0.006
Male ≥8	18 (22)	6 (12)	12 (39)
Female <4	35 (64)	31 (79)	4 (25)	0.0001
Female ≥4	20 (36)	8 (21)	12 (75)
Depressive symptoms (CES-D ≥ 16)	83 (61)	56 (64)	27 (57)	0.48
CD4 cell count/mm3, median IQR (n=131)	521 (343, 796)	532 (374, 809)	468 (303, 745)	0.19
Antiretroviral Therapy	131 (97)	85 (97)	46 (98)	0.68
HIV RNA < 50 copies/mL	107 (79)	69 (78)	38 (81)	0.73
HCV Genotype 1a	107 (79)	70 (80)	37 (79)	0.91
Liver Stiffness (Elastography) kPa (n=128)
≤ 8	84 (66)	58 (71)	26 (57)	0.27
8.1–11.9	29 (23)	16 (20)	13 (28)
≥ 12	15 (12)	8 (10)	7 (15)

^*9 individuals missing PEth data in CHAMPS study (N=144 CHAMPS study)

HCV: Hepatitis C virus

UC: Usual Care

IQR: Interquartile range

CES-D: Center for Epidemiologic Studies Depression Scale (Range 0 to 60)

HIV: Human immunodeficiency virus

In univariable analysis, only current use of heroin/cocaine (urine toxicology) was associated with recent heavy alcohol use (PEth ≥ 50 ng/ml) (Relative Risk [RR] 1.84, 95% CI 1.12–3.01) (see supplemental table). In a multivariable model including study group, age, sex, race, employment status, urine toxicology, liver disease stage, and CES-D, a positive urine toxicology for heroin/cocaine (RR 1.65, 95% CI 1.05–2.59) remained significantly associated with PEth ≥ 50 ng/ml (see supplemental table). In the multivariable model, participants in the UC+Incentive intervention group (RR 0.56, 95% CI 0.33–0.96) were less likely to have a PEth ≥ 50 ng/ml.

LDV/SOF was initiated in 103 of 135 participants and, of those who initiated LDV/SOF, SVR was achieved in 92% (Figure 1). In univariable analysis, no factors were significantly associated with failure to initiate treatment. Neither an AUDIT score suggestive of hazardous/heavy drinking (male≥8; female ≥4) (RR 1.04; 95% CI 0.47–2.30) nor PEth ≥ 50 ng/ml (RR 0.72, 95% CI 0.35–1.48) was significantly associated with failure to initiate HCV treatment. In a separate multivariable model which defined heavy alcohol as either an AUDIT ≥8 in males/≥4 in females or PEth ≥ 50 ng/ml, heavy alcohol use (RR 0.93, 95% CI 0.45–1.92) was not significantly associated with failure to initiate HCV treatment (see supplemental table). Only male sex was significantly associated with failure to initiate treatment (adjusted RR 2.33, 95% CI 1.08–5.00) in the multivariable model with hazardous alcohol use defined by PEth (Table 2).

Figure 1-.

HCV TREATMENT OUTCOMES BY PETH (50 < or ≥ 50 ng/ml)

LDV/SOF: Ledipasvir/Sofosbuvir

SVR: Sustained Virologic Response

Table 2-.

RELATIONSHIP BETWEEN HEAVY ALCOHOL USE (AUDIT AND PETH) AND FAILURE TO INITIATE HCV TREATMENT AND ACHIEVE CURE

	Failure to initiate treatment RR (95% CI)			Failure to achieve HCV cure (including failure to start and treatment failure) RR (95% CI)			Failure to achieve HCV cure (sensitivity analysis including only individuals who initiated treatment and were treatment failures) RR (95% CI)
	Unadjusted	Adjusted (AUDIT model)	Adjusted (PEth Model)	Unadjusted	Adjusted (AUDIT model)	Adjusted (PEth model)	Unadjusted	Adjusted (AUDIT model)	Adjusted (PEth model)
Randomization (Reference: UC)
UC	REF	REF	REF	REF	REF	REF	REF	REF	REF
UC+Peer	0.57 (0.26, 1.26)	0.47 (0.21, 1.06)	0.43 (0.19, 0.97)	0.69 (0.36, 1.32)	0.58 (0.29, 1.19)	0.57 (0.28, 1.16)	1.07 (0.21, 5.41)	0.61 (0.30, 1.27)	0.62 (0.31, 1.28)
UC+Incentives	0.86 (0.43, 1.72)	0.69 (0.32, 1.47)	0.64 (0.30, 1.35)	0.88 (0.48, 1.61)	0.75 (0.38, 1.47)	0.73 (0.38, 1.41)	0.93 (0.17, 5.16)	0.78 (0.39, 1.59)	0.82 (0.41, 1.63)
Age ≥ 55	0.95 (0.52, 1.74)	0.83 (0.43, 1.62)	0.80 (0.41, 1.55)	1.13 (0.68, 1.88)	0.87 (0.49, 1.53)	0.84 (0.48, 1.48)	2.12 (0.56, 8.02)	0.85 (0.47, 1.53)	0.84 (0.47, 1.51)
Male	1.52 (0.78, 2.94)	2.22 (1.02, 4.76)	2.33 (1.08, 5.00)	1.47 (0.85, 2.56)	1.89 (0.97, 3.70)	1.92 (1.00, 3.70)	1.56 (0.41, 5.88)	1.72 (0.91, 3.33)	1.69 (0.89, 3.13)
Non-Black Race	1.17 (0.42, 3.22)	1.58 (0.48, 5.25)	1.50 (0.48, 4.70)	0.89 (0.33, 2.42)	1.21 (0.40, 3.64)	1.20 (0.42, 3.44)	--	--	--
Employed	0.41 (0.11, 1.56)	0.37 (0.09, 1.55)	0.32 (0.08, 1.32)	0.31 (0.08, 1.19)	0.32 (0.08, 1.27)	0.29 (0.07, 1.17)	--	--	--
Positive Urine for Cocaine/Heroin	1.67 (0.88, 3.18)	1.43 (0.75, 2.71)	1.54 (0.80, 2.96)	1.18 (0.68, 2.03)	1.05 (0.59, 1.85)	1.09 (0.61, 1.95)	0.22 (0.03, 1.74)	1.07 (0.61, 1.87)	1.07 (0.60, 1.93)
Depressive symptoms CES-D ≥16	1.38 (0.71, 2.67)	1.24 (0.60, 2.54)	1.18 (0.57, 2.47)	1.09 (0.64, 1.85)	1.11 (0.61, 2.01)	1.09 (0.59, 1.99)	0.55 (0.16, 1.93)	1.11 (0.60, 2.06)	1.12 (0.60, 2.11)
Liver Stiffness (Reference: ≤ 8kPa)
≤ 8.0 kPa	REF	REF	REF	REF	REF	REF	REF	REF	REF
8.1–11.9 kPa	1.63 (0.81, 3.28)	1.29 (0.63, 2.66)	1.34 (0.64, 2.79)	1.74 (0.98, 3.10)	1.36 (0.71, 2.57)	1.39 (0.72, 2.65)	2.55 (0.62, 10.46)	1.42 (0.74, 2.74)	1.42 (0.73, 2.78)
≥12 kPa	1.75 (0.76, 4.06)	1.87 (0.78, 4.48)	2.09 (0.88, 4.97)	1.68 (0.81, 3.48)	1.80 (0.86, 3.74)	1.95 (0.93, 4.11)	1.70 (0.21, 13.72)	1.65 (0.77, 3.50)	1.68 (0.77, 3.65)
AUDIT (Male ≥8; Female ≥4)	1.16 (0.61, 2.22)	1.04 (0.47, 2.30)		1.06 (0.60, 1.85)	1.11 (0.57, 2.17)		0.77 (0.17, 3.47)	1.23 (0.67, 2.27)
PEth (≥50 ng/ml)	1.12 (0.60, 2.09)		0.72 (0.35, 1.48)	1.08 (0.64, 1.83)		0.85 (0.46, 1.57)	0.97 (0.26, 3.65)		1.06 (0.58, 1.93)

HCV: Hepatitis C virus

UC: Usual Care

CES-D: Center for Epidemiologic Studies Depression Scale (Range 0 to 60)

In univariable analysis, no factors were associated with failure to achieve HCV cure (which included individuals failing to initiate treatment and those who initiated treatment but did not achieve SVR) (Table 2). Neither an AUDIT score suggestive of hazardous/heavy drinking (male≥8; female ≥4) (RR 1.11; 95% CI 0.57–2.17) nor PEth ≥ 50 ng/ml (RR 0.85, 95% CI 0.46–1.57) was significantly associated with failure to achieve cure. In a separate multivariable model which included heavy alcohol use by either AUDIT or PEth (AUDIT male≥8; female ≥4 or PEth ≥ 50 ng/ml), heavy alcohol (RR 1.10, 95% CI 0.59–2.04) was not significantly associated with failure to achieve cure (see supplemental table). A sensitivity analysis limited to only participants that initiated treatment (N=103) revealed that neither an AUDIT score suggestive of hazardous/heavy drinking (male≥8; female ≥4) (RR 1.23; 95% CI 0.67–2.27) nor PEth ≥ 50 ng/ml (RR 1.06, 95% CI 0.58–1.93) was significantly associated with failure to achieve cure (Table 2). No other characteristics were associated with failure to achieve cure in the univariable or multivariable models (Table 2).

Discussion

In this clinic-based research cohort of HIV/HCV co-infected persons who use drugs, heavy alcohol use was common by AUDIT (28% hazardous/heavy drinking) and blood PEth (35% with PEth ≥ 50 ng/ml) and if either AUDIT or PEth is taken as an indication of heavy alcohol use then 61 (45%) of individuals meet criteria. While these alcohol assessment methods, self-report and biomarker, did have considerable overlap in detecting heavy alcohol use, they also had significant differences in detection. Of the 47 individuals with a PEth ≥ 50 ng/ml, 23 (49%) reported no or minimal drinking by AUDIT. It is concerning that in this cohort of HIV/HCV co-infected persons who use drugs heavy alcohol use was frequently not detected by self-report (AUDIT) as alcohol use can lead to increased fibrosis and may contribute to progressive liver disease after HCV cure.^2,3,7,17 On the other hand, it was encouraging that heavy alcohol use either by self-report or biomarker was not associated with failure to initiate HCV treatment or failure to achieve HCV cure. This finding strongly supports the recommendations from the HCV Guidance Panel which prioritize HCV treatment for persons who actively use alcohol including those with alcohol use disorders in whom elimination of HCV will remove one liver threat.¹³

In our study, participants with a PEth ≥ 50 ng/ml were significantly more likely to have urine toxicology positive for cocaine and heroin. This reinforces findings from other research studies that have reported high rates of co-occurring substance use disorders and specifically high rates of alcohol use among persons using drugs.^25–28 These data suggest the need to co-locate HCV, alcohol, and other drug treatment in order to make care easily accessible. Additionally, in reference to substance use, there has often been a concern among medical professionals that substance use could impact HCV treatment adherence and cure rates. In the era of interferon, heavy alcohol consumption was associated with higher discontinuation rates and poorer treatment responses.^8,10,29 In fact, alcohol impaired the immunological responses and reduced the effectiveness of interferon and thus alcohol cessation was often required.^8,29 However, DAAs require a shorter duration to achieve cure, and their mechanisms of action are not biologically affected by alcohol.

Despite this, some providers and insurance plans still require alcohol cessation due to continued fears about treatment adherence and the effectiveness of treatment.^14,16 Our study shows that in this cohort of HIV/HCV co-infected persons who use drugs with heavy alcohol use, individuals were still able to initiate DAA-based HCV treatment and achieve SVR/cure. Hence, clinicians and insurers should not restrict access to HCV treatment in the era of DAAs based on alcohol use but instead provide increased support during treatment based on the patient’s individual needs. This is extremely important as individuals with HCV infection who have HIV coinfection and/or use alcohol are at elevated risk of liver related death. An analysis of US multiple-cause mortality data from 1999–2007 from the CDC revealed significant increases in HCV related deaths and that both HIV and alcohol were factors associated with HCV-related deaths.³⁰ HCV cure removes one of these important insults to the liver.

The documented high rates of alcohol use in this study also underscore the need to counsel patients around alcohol use and liver disease throughout the HCV treatment process. HCV treatment providers can play an important role in alcohol counseling by incorporating alcohol brief interventions into HCV treatment and referring individuals with alcohol use disorders to treatment programs to try to avoid continued damage to the liver.^31–33 Ongoing alcohol use post HCV treatment has the potential to offset the gains achieved by HCV treatment and lead to continued fibrosis of the liver. For example, research has documented that heavy alcohol use is associated with the highest incidence of end-stage liver disease and liver-related death irrespective of whether patients have recovered from HCV infection or if they remain chronically infected.³⁴ This is particularly relevant to persons with significant liver fibrosis who achieve HCV cure since persistent, heavy alcohol use may lead to progressive liver injury. Indeed, in the era on HCV DAAs, alcohol use disorder has emerged as one of the leading indications for liver transplantation.

Our study also highlights the limitations of alcohol ascertainment by self-report (AUDIT) in persons with HCV disease. Notably, of the 47 individuals with a PEth ≥ 50 ng/ml at baseline (which measured alcohol levels over the most recent 2–3 weeks), almost half (49%) reported drinking never, monthly or less, or 2–4 times per month by AUDIT.^19,21,24 The PEth ≥ 50 ng/ml is suggestive of daily drinking of two or more drinks per day which does not align with the AUDIT (self-report) data for almost half the individuals in this category.^21,24 While other studies have suggested a high correlation between AUDIT and PEth, individuals in our study may have believed that reporting high alcohol use would impact their HCV clinical care.^35,36 Some patients, particularly those with longstanding HCV infection, are aware that alcohol cessation was often required with interferon and that it is still required by some clinicians and insurance plans. This knowledge may have driven some patients to underreport their alcohol consumption. It is also possible that since the AUDIT inquired about alcohol use over the preceding 12 month period and the PEth noted alcohol use in the most recent 2–3 weeks that some patients just reported this information in a different manner or noted their most common practice over 12 months which was different than their prior 2–3 week period.

There are several limitations to this study. First, this study was based within a clinic and may not fully apply to community-settings where participants must first make it to the clinic and overcome the barrier of linkage to care. Second, each participant in this study received treatment adherence support based on evaluation by their clinician. Clinical sites would need to develop a protocol to ensure that patients received the needed support based on alcohol use or other barriers encountered during treatment. Finally, these data were collected within the context of a randomized trial where most participants received some sort of intervention. It will be important to document these findings in real-world settings. This study is novel as it is the first study that objectively documented the impact of heavy alcohol use through PEth on the HCV care continuum as compared to other studies that relied on self-report through AUDIT.³⁷

In summary, this research suggests that in the era of DAAs, individuals with alcohol use including those with heavy alcohol use are able to initiate HCV treatment and achieve cure. For these individuals at high risk of liver fibrosis, cirrhosis and death, curative HCV treatment may offer a reduction in morbidity and mortality. Hopefully, this study and similar research in this area will continue to open up access to DAAs to those with alcohol use. Treatment for hepatitis C offers an opportunity for clinicians to engage patients in alcohol counseling pre/post HCV treatment which can put them on a journey to continued liver health.